Three relatively new models for predicting risk of mortality and dependency may ultimately help guide treatment decisions and improve outcomes in patients with newly diagnosed Parkinson disease (PD), according to a study published in Movement Disorders.
Investigators sought to establish clinically valid models that could predict risk for mortality, dependency, and dependency or death in patients who have received a recent PD diagnosis. Models were formed in the Parkinsonism Incidence in North-East Scotland study and were previously validated in the ParkWest study, which involved 192 patients in Norway.
In the mortality model, the investigators identified older age, male gender, Charlson comorbidity index, and increased severity of axial features as independent predictors for death. In addition, the researchers found that a greater history of smoking, a lower Mini-Mental State Examination score, increased severity of axial features, and increasing age were all independent predictors for dependency and death or dependency in newly diagnosed patients with PD.
The models featured very good or good discrimination; however, the investigators believe that greater improvement in the models’ precision is necessary for identifying individualized risk prediction. As such, the investigators suggest recalibration of the models prior to attempting to stratify risk at the individual level.
Since these models were developed and utilized in the United Kingdom, the investigators also suggest recalibration of the models for specific geographic locations. In addition, the investigators state that the study’s findings are limited by the small sample size.
Prognostic risk models featured in this study may also provide clinical utility for stratifying treatment “if disease-modifying treatments with potentially serious side effects become available for PD.”
Macleod AD, Dalen I, Tysnes OB, Larsen JP, Counsell CE. Development and validation of prognostic survival models in newly diagnosed Parkinson’s disease [published online October 4, 2017]. Mov Disord. doi:10.1002/mds.27177